What the bookkeeping model actually separated
A team writing in Nature Geoscience rebuilt China's forest carbon balance from 1986 to 2019, fitting aboveground biomass accumulation curves locally for each 1-degree grid cell and running a spatially explicit bookkeeping model over forest maps at 30-metre resolution. The biomass sink averages 0.139 ± 0.052 petagrams of carbon a year across the period, and it grows within it: 0.1 ± 0.015 over 1990 to 1999, 0.2 ± 0.012 over 2010 to 2019. What the model adds to that total is a split — uptake on land that had never carried forest, tracked separately from uptake on land that had carried it and lost it.[1]
The two lines move in opposite directions. Afforested tree cover expanded faster, 1.31 million hectares a year against 1.06 million for reforestation after disturbance. The carbon went the other way: regrowth accumulated 1.47 ± 0.42 tonnes per hectare a year, afforestation 0.96 ± 0.28. Over three decades the slower-growing, faster-expanding category loses to the faster-growing, slower-expanding one, and reforestation ends up the dominant contributor to the national sink.[1]
The gap in growth rate is the part worth holding onto, because it is a per-hectare quantity and policy is written in hectares. A site that has grown forest before generally keeps the soil, the seed bank, the mycorrhizal network and the microclimate that made it possible; a bare hillside planted for the first time has to build those before it can accumulate quickly. Other readings stay open. Stands regrowing after disturbance may sit disproportionately on better ground to begin with, so part of the 1.47 may come from site quality rather than from regrowth as a process, and the study's grid-cell curves cannot fully separate the two.[1]
The same quantity, squeezed from the other end
A second paper published the same day works on the same per-hectare quantity from the other end. Using a global database of tree-ring isotopes, eddy-covariance fluxes and satellite vegetation metrics, it reports that fine particulate pollution has a predominantly negative effect on plant water-use efficiency, and that the decline comes primarily from suppressed photosynthesis rather than from changed evapotranspiration: less photosynthetically active radiation reaching the leaf, lower carboxylation capacity inside it.[2]
Put beside each other, the two results say the same thing about accounting. The mapped area of a forest is the easy variable to report and the one a programme can be scored on; the rate at which a hectare fixes carbon is the variable that decides the sink, and it is set by things the map does not carry — whether the ground has grown forest before, and what the air above it is carrying. A national inventory that counts hectares well and rates poorly will misplace credit in both directions at once.[1], [2]
The aerosol paper adds one detail that matters for anyone using model output as a baseline: ecosystem models that omit aerosol processes cannot reproduce the observed relation at all, and covarying climate drivers alone do not account for it. On this evidence, whatever a model says about future carbon uptake in a polluted region is missing a term.[2]
Which number the target should be written in
Three days ago this column looked at 103 European cities aiming at net zero: they had booked 61.4 million tonnes of residual emissions for compensation while only 18 per cent of that volume had a quantified removal behind it. The Nature Geoscience reconstruction is the same ledger with the number filled in, and filling it in moves the credit; this time China supplies the number. Once uptake is measured per hectare by stand origin, protecting forest that is already regrowing after disturbance does more per hectare than adding the next plantation, which is the opposite of how an area target reads.[1], [3]
The authors put the practical form of this plainly: prioritise protecting regrowing forests alongside targeted afforestation. For a country aiming at carbon neutrality, the observable signal is narrow and checkable — whether forest carbon accounting starts reporting biomass accumulation separately for post-disturbance regrowth and for new plantations. Until it does, an expanding planted area and a strengthening sink will keep being read as the same fact.[1]